Mixed Chimerism at Three Months after Allogeneic Hematopoietic Stem Cell Transplantation Is a Significant Predictor of Disease Relapse in High-Risk Acute Myeloid Leukemia Patients in First Complete Remission

Background:

Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is a potential curative strategy for high-risk acute myeloid leukemia (AML) patients. However, the risk for disease recurrence following allo-HSCT remains significant and associated with poor outcomes. The ability to predict relapse before detectable morphologic recurrence may allow for preemptive interventions, such as immune modulation, donor lymphocyte infusion (DLI), or initiation of hypomethylating agents, to potentially augment graft-versus-leukemia effect. Post-transplantation peripheral blood chimerism analysis, represents a potential tool to predict disease recurrence, although a validated consensus on the use of this technique in the post-allo-HSCT follow-up has not been established yet and its precise role in this setting remains unclear. The aim of our study is to evaluate the impact of chimerism evaluation 3 months after allo-HSCT in AML patients who were in first complete remission (CR1) before transplantation and remained in clinical and morphological CR at day 90 post-transplantation, on overall survival (OS) and relapse incidence.

Patients and methods:

We evaluated 194 AML patients who received allo-HSCT at our center between January 2006 and December 2014 and for whom chimerism follow-up has been performed at 3 months. There were 103 (53%) males and 91 (47%) females with a median age of 43 years (range: 18-67). Patients were classified according to the European LeukemiaNet classification for cytogenetic and molecular biology markers (Dohner et al. Blood 2010), accordingly, 64% patients were unfavorable and 36% were in intermediate II risk group. At allo-HSCT, all patients were in CR1; 136 (70%) received a full intensity conditioning (MAC) and 58 (30%) received a reduced intensity conditioning (RIC). HSC donors were 72 (37%) HLA identical siblings (44 BM and 36 PBSC), 54 (28%) 10/10 HLA matched unrelated donors (35 BM and 19 PBSC), 33 (17%) 9/10 HLA mismatched unrelated donors (16 BM and 17 PBSC) and 35 (18%) double cord blood units (only 7 were 6/6 HLA matched). Chimerism analysis was performed on marrow and/or blood samples every month following allo-HSCT using polymerase chain reaction (PCR) based on informative polymorphic short tandem repeat, a mixed chimerism was defined by having 5% or more of recipient cells. 

Results:

At day 90 after transplantation, all patients remained in clinical and cytological CR at time of chimerism evaluation, 155 (80%) had a full donor chimersim (FDC) and 39 (20%) had a mixed chimerism (MC) ranging from 65% to 95% of donor cells. Among patients with MC, 9 (23%) received increasing doses of DLI (5 of them reached FDC at 6 months), while 20 (51%) could not receive DLI (7 because transplanted from cord blood, and 13 because of the presence of GVHD), the rest of patients were left with a transient MC and regained FDC during follow-up.

After a median follow-up of 34 months (range: 4-96) for the surviving patients, the median OS in patients with FDC was not reached with a 3 years probability of 62% (95% CI: 58-66), and for patients with MC, it was 18 months (12-24) with a 3 years probability of 32% (95% CI: 23-41), (p=0.01). Twenty-two patients in the MC group have progressed during the follow-up and 17 among them died from disease progression. The cumulative incidence of relapse at 3 years was 25% (95% CI: 21-29) for FDC patients and it was 70% (95% CI: 62-80) for MC patients, (p<0.001). The impact of mixed chimerism was still valid in multivariate analysis after including patient age, type of donor, HSC source and risk group, and was independent from the intensity of the conditioning regimen with a Hazard Ratio of 4.7, and a 95% CI: 2.6-8.4, p<0.001.

Conclusion:

We demonstrated that chimerism evaluation at day 90 after allo-HSCT is an independent predictor of disease relapse in patients who remain in CR at that time and significantly impacts on long term survival. The standardization of this evaluation may lead to the identification of patients with high-risk of relapse risk suggesting the need of early preemptive intervention.